Research

  Published Work

 

 

 

 

  • Over-Education in Multilingual Economies: Evidence from Catalonia pdficon

Joint with Maite Blázquez ( Universidad Autónoma de Madrid).

Forthcoming at International Migration.

Abstract: Individuals with deficient language skills may compensate for this disadvantage in the labor market by acquiring more formal skills. Catalonia's economy is characterized by linguistic diversity and provides thereby a unique opportunity to measure the incidence of language proficiency on over-education. Catalan language, formerly confined to informal uses, became co-official with Spanish and the language of instruction in the early eighties. This change, however, did not undermine the intensive use of Castilian in most spheres of communication. Descriptive evidence seems to suggest that individuals with better language knowledge are more likely to be over-educated. This can lead us to think, as is usually the case in the public discussions, that individuals who are not fluent in the language of instruction tend to under-educate, since they are discouraged to attend school. However, once we estimate a model that controls for individuals’ socio-demographic characteristics, the opposite emerges: language knowledge is shown to have in fact a negative effect on over-education. This effect, although robust to accounting for endogeneity of language knowledge and significant at the individual level, is mostly non-significant on average.

Non-Tobin’s q in Tests for Financial Constraints to Investment pdficon

Book Chapter forthcoming in “The Economics of Imperfect Markets,” Springer Book Series “Contributions to Economics.”

Abstract: Liquidity constrained firms may be under two very well identified investment regimes, constrained and unconstrained. In this paper I derive theoretical investment equations for both regimes and discuss the consequences of ignoring the specific form of the liquidity constrained regime. I also show that expressing the investment equation as a function of Tobin’s q is by no means necessary in theory and in practice, in particular, it is not required to test for liquidity constraints.

  • International Job Search: Mexicans in and out of the U.S. pdficon (also a WP at ITAM, IZA and Stony Brook)

Joint with Alfredo Cuecuecha (ITAM).

Review of Economics of the Household 7(4), 2009.

Abstract:   It is argued that migration from Mexico to the US and its corresponding return migration are determined by international wage differentials and preferences for origin. We use a model of job search, savings and migration to show that job turnover is a crucial determinant of the migration process. We estimate this model by Simulated Method of Moments (SMM) and .find that migration practically disappears if Mexico has American arrival rates while employed. Doubling migration costs reduces migration rates in half, while subsidizing return migration in $300 reduces migration rates of older migrants but increases migration rates of younger migrants.

  • Employment and Deadweight Loss Effects of Observed Non-Wage Labor Costs pdficon (also a WP at ITAM, IZA and PUCP)

Joint with Giovanna Aguilar (Universidad Católica, Lima).

Economic Inquiry 48(1), 2010.

Abstract:   To assess the employment effects of labor costs it is crucial to have reliable estimates of the labor cost elasticity of labor demand. Using a matched firm-worker dataset, we estimate a long run unconditional labor demand function, exploiting information on workers to correct for endogeneity in the determination of wages. We evaluate the employment and deadweight loss effects of observed employers' contributions imposed by labor laws (health insurance, training, and taxes) as well as of observed workers' deductions (social security, and income tax). We find that non-wage labor costs reduce employment by 17% for white-collars and by 53% for blue-collars, with associated deadweight losses of 10% and 35% of total contributions, respectively. Since most firms undercomply with mandated employers' and workers contributions, we find that full compliance would imply employment losses of 4% for white-collars and 12% for blue-collars, with respective associated deadweight losses of 2% and 6%.

  • Matching Bias in Labor Demand Estimation pdficon (also a WP at IZA)

Economic Letters 100, pp. 297/299, 2008.

Joint with Giovanna Aguilar (Universidad Católica, Lima).

Abstract:   Using a matched .rm-worker dataset, we show both theoretically and empirically that positive assortative matching between .rms and workers leads to an underestimation of the absolute value of wage elasticity of labor demand.

  • Does Wealth Explain Black-White Differences in Early Employment Careers? pdficon

Journal of Business and Economic Statistics 25(4) pp. 484-500 (2007).

Abstract: In this paper I inquire on the effects initial wealth has in black-white differences in early employment careers. I set up a dynamic model in which individuals simultaneously search for a job and accumulate wealth, and fit it to data from the National Longitudinal Survey (1979-cohort). The estimates show that borrowing constraints are tight for both race groups. Regime changes reveal that initial wealth accounts fully for the racial gap in wealth and wages at the beginning of employment careers. When growth in wage and arrival rates is also taken into consideration, it also accounts for around 76% of the wealth gap and 90% of the wage gap persisting several years after graduation. By contrast, labor market variables are shown to explain 100% of the gap in wealth and wages persisting several years after High School graduation. 

  • The Catalan Premium: Work and Language in Catalonia. pdficon

Journal of Population Economics 20(3) pp. 669-686 (2007).

Abstract: In this paper I measure the contribution of knowing Catalan to finding a job in Catalonia. In the early eighties a drastic language policy change (normalització) promoted the learning and use of Catalan in Catalonia and managed to reverse the falling trend of its relative use versus Castilian (Spanish). Using census data for 1991 and 1996, I estimate a significant positive Catalan premium: the probability of being employed increases between 3 and 5 percentage points if individuals know how to read and speak Catalan; it increases between 2 and 6 percentage points for writing Catalan.

  • Job Search and Asset Accumulation under Borrowing Constraints. pdficon

International Economic Review, 47(1) pp.233-263 (2006).

Abstract: This paper examines the relationship between wealth accumulation and job search dynamics. It proposes a model in which risk-averse individuals search for jobs, save into a risk-free asset, and borrow to smooth their consumption. In the model, one motivation for accumulating wealth is to finance voluntary quits in order to search for a better job. Using data on men from the National Longitudinal Survey (1979 cohort), I estimate by maximum likelihood the individual’s dynamic decision problem and use the resulting optimal policies to simulate joint employment and saving histories. The results show that borrowing constraints are tight and reinforce the influence of wealth on job acceptance decisions, namely that more initial wealth and access to larger amounts of credit increase wages and unemployment duration.

Firm Investment in Imperfect Capital Markets: A Structural Estimation. pdficon

Joint with Sangeeta Pratap (ITAM).

Review of Economic Dynamics 6(3) pp. 513-545 (2003).

Abstract: In this paper we characterize and estimate the degree to which liquidity constraints affect real activity. We set up a dynamic model of firm investment and debt in which liquidity constraints enter explicitly into the firm's maximization problem, so that investment depends positively on the firm's financial position. The optimal policy rules are incorporated into a maximum likelihood procedure to estimate the structural parameters of the model. We identify liquidity constraints from the dynamics of a firm's evolution, as formalized by the dynamic estimation process, and find that they significantly affect investment decisions of firms. Firms ability to raise equity is about 73% of what it would have been under free capital markets. If firms can finance investment by issuing fresh equity, rather than with internal funds or debt, average capital stock is about 6% higher over a period of 20 years. Transitory interest rate shocks have a sustained impact on capital accumulation, which lasts for several periods.

   Working Papers and Papers under Review

 

 

 

  • Job Creation and Investment in Imperfect Capital and Labor Markets.pdficon

Abstract: This paper shows that liquidity constraints restrict job creation even with flexible labor markets. In a dynamic model of firm investment and demand for labor with imperfect capital markets, represented as a constraint on dividends, and imperfect labor markets, contained in legal firing costs applicable to some workers, firms use flexible labor contracts to alleviate financial constraints. The optimal policy rules of the theoretical model are integrated into a maximum likelihood procedure to recover the model's behavioral parameters. Data for the estimation come from the CBBE (Balance Sheet data from the Bank of Spain). I evaluate the effects of removing one imperfection at a time, and show that the relaxation of financial constraints produces

(i)                 more job creation than the elimination of labor market rigidities, and

(ii)               a substantial increase in firm investment, which does not happen if only labor market rigidities are removed.

  • Unemployment Dynamics and Social Security. 

Abstract: The goal of this paper is to determine the effects of different social security regimes on job search. On the one hand, a less generous pension system induces higher savings across the life cycle and makes wealthier agents more reluctant to accept low wage offers. On the other hand, as the social security system provides insurance against labor shocks, such as layoffs, a less generous system induces agents to accept bad job offers to save for retirement. To determine the strength of each effect, we develop a quantitatively life-cycle overlapping generations model with search, savings and a balanced-budget social security. After obtaining the parameters that match the predicted moments to the observed path of employment states, wages, and assets, we compute the effects of alternative social security regimes in the U.S. economy.

  • Family Job Search and Consumption. 

Joint with Ignacio García-Pérez (Universidad Pablo de Olavide).

Abstract: In this paper we find out the determinants of consumption variations depending on the employment transitions experienced by household members. We set up a utility-maximizing household search model in which consumption and job search decisions are made jointly. Families determine a level of consumption, who has to work, and what is the minimum acceptable wage for each family member. This interaction implies that each member's reservation wage is highly dependent not only on the partner's labor market status but also on his/her wage. In this model, not only wealth but also the employment status of the partner allows agents to be more selective and search longer. Moreover, we show that reservation wages describe an inverted U in the partner's wage. This effect, which we name "the new added worker effect", is more important the higher is the wealth of the household.  Using the Spanish Continuous Family Expenditure Survey (ECPF) for the period 1985-1996, we estimate the behavioral parameters of the theoretical model for a sample of Spanish households. With the results of this estimation, we simulate some policy experiments. In particular we are interested on how the Unemployment Benefits system affects intra-household decisions. 

  • Remittances, Savings and Human Capital working icon

Joint with Alfredo Cuecuecha (ITAM).

Abstract:   Recent literature has shown that migration and remittances can affect inequality in communities through the accumulation of wealth, and that migration and remittances have an ambiguous sign on its relation with human capital accumulation. Disentangling these two effects is difficult because while migration and remittances have an effect on asset and human capital accumulation, the stocks of these variables can also affect levels of migration and remittances. These paper presents a model in which households decide optimally these variables in a dynamic setting. Preliminary evidence shows that migration and remittances have different effects on the accumulation of assets and human capital. These effects indicate that short run effects and long run effects may have different signs.

  • Great Depressions and Fake Recoveries: Peru, 1980-2000working icon

Joint with Carlos Urrutia (ITAM).

Abstract:    In 1982, the Peruvian economy faced two large negative shocks: (i) an unprecedented natural disaster, and (ii) a radical worsening on international conditions (the “Debt Crisis'”). Other Latin American countries (as Mexico and Chile) suffered similar negative shocks in the same year, leading to a region-wide depression. Bergoeing et al (2002) document that the recovery in Chile took less than five years, while in Mexico output remained below trend for about fifteen years. Peruvian response to the 1982 crisis is very different to both cases. In 1985, Alan Garcia took power in Peru and started a large scale, demand-driven, recovery package. Expansionary fiscal and monetary policies led to a surprisingly fast recovery between 1985 and 1987. However, the last two years of Garcia's mandate were signed by an even deeper recession, together with a collapse of the currency system (hyperinflation). A very painful recovery started with the next government and took all the nineties.
In this paper, we analyze the Peruvian experience through the lens of Neoclassical Growth Theory. We show that the 1982 crisis looks like a negative TFP shock. We observe that TFP didn't recover substantially during Garcia's expansion, but remained stagnant, and fell again at the end of his mandate. The whole recovery attempt was successful at increasing inputs (capital, labor), but failed at increasing the efficiency in the use of these inputs. Moreover, the collapse in the exchange system due to the expansion in money supply had ultimately negative real effects on TFP. That is why we label Garcia's experiment a “fake recovery'”. Using an extended version of the Neoclassical Growth model, we are able to account for both the short-lived expansion between 1985 and 1987, the disaster between 1988 and 1990, and the further - but slow - recovery of the Peruvian economy.

  • Job Search and Marital Dynamics working icon

Abstract:    This paper explains jointly marital and employment decisions with a search-theoretic framework. Agents of different sex meet and decide to get married. Simultaneously, they can receive job offers that they are free to accept or reject. Employment transitions are crucially influenced by their marital status, and reciprocally, marriage and divorce are determined by employment status. The model predicts comovements between labor market and marital transitions, and marriages of agents of similar employment status and wage levels. The model is fitted to data from the National Longitudinal Survey - Youth Cohort and used to analyze the effect of changes in the labor market environment, such as a higher unemployment rate, on the formation and dissolution of households.

  • Diffusion of Knowledge in New Industries.pdficon

Abstract: This paper shows firms' entry and exit in new industries as a process of Bayesian learning about their types, as in Jovanovic (1982), but also about the profitability of the new industry. Not only incumbents learn about the new industry, but also prospective entrants. Entry and exit is a function of this twofold learning. It is shown that: (i) there is a feedback loop between number of firms and cost reduction; (ii) learning curves rise more and more sharply for later entrants, because they face less uncertainty about the value of the industry; (iii) there is a shake-out, a reduction in the number of firms, when uncertainty about the industry declines and the most efficient firms expand their production. 

  • Informational Matching. pdficon

Abstract: This paper analyzes the problem of matching heterogeneous agents in a Bayesian learning model. One agent gives a noisy signal to another agent, who is responsible for learning. If production has a strong informational component, a phase of cross-matching occurs, so that agents of low knowledge catch up with agents of higher knowledge. It is shown that

(i) a greater informational component in production makes cross-matching more likely;

(ii) as a new technology is mastered, production becomes relatively more physical and less informational;

(iii) greater dispersions of the ability to learn and transfer information make self-matching more likely; and

(iv) higher inequality leads to self-matching and to more inequality, whereas lower inequality leads to cross-matching, which can make less productive agents not only catch up with but also overtake more productive ones. 

  • Fixed and Random Effects in Classical and Bayesian Regression.pdficon

Abstract: This paper proposes a common and tractable framework for analyzing different definitions of fixed and random effects in a constant-slope variable-intercept model. It is shown that, regardless of whether effects (i) are treated as parameters or as an error term, (ii) are estimated in different stages of a hierarchical model, or whether (iii) correlation between effects and regressors is allowed, when the same information on effects is introduced into all estimation methods, the resulting slope estimator is also the same across methods. If different methods produce different results, it is ultimately because different information is being used for each method. 

  • Capital Adjustment Costs that mimic Liquidity Constraints in a Two Period Model.pdficon

Abstract: In a two-period deterministic model with capital as the only factor, adjustment cost functions induce liquidity constrained investment when they reproduce the shape of the shadow price of financial resources. 

  • Working and Studying in Rural Latin American: Critical Decisions of Adolescence.pdficon

Abstract: This paper evaluates the determinants of school attendance and work of rural adolescents between 10 and 18 years old in 1997-1998 for a sample of Latin American countries. Rural adolescents are quite disadvantaged relative to their urban counterparts. The share of rural adolescents studying while concurrently working part-time is significantly higher, household income is significantly lower, “supply-side” issues are an important factor in rural non-attendance, and to the extent that the educational attainment of the parents creates inter-generational persistence we find that rural youth are starting from a disadvantaged position. We present some statistical analysis that highlights these problems and also perform bivariate binary estimation to identify the determinants of these decisions. We find that for most countries critical determinants for making these choices are household income and parental education as well as household composition. Further, we find that there is evidence of a significant “trade-off” between working and studying. Finally, inter-generational factors allow for both a virtuous cycle and a vicious cycle.

  • Job Creation under Liquidity Constraints: The Spanish Case. pdficon

Abstract: This paper shows that liquidity constraints restrict job creation even with flexible labor markets. In a dynamic model of labor demand, I show that under imperfect capital and imperfect labor markets, firms use temporary contracts to relax financial constraints. Evidence for the predictions of the model is presented using Spanish data from the CBBE (Balance Sheet data from the Bank of Spain). The creation of temporary contracts in 1984 implied an insufficient alleviation of liquidity constraints; consequently, permanent job creation remains slow. A job creation strategy should not only remove labor rigidities, but also financial constraints.

  • Occupational Selection in Multilingual Labor Markets: The Case of Catalonia. pdficon 

Joint with Núria Quella (Stony Brook University).

Abstract: Agents with high proficiency in languages may be selected into occupations that require high communication skills, e.g., occupations in which they must face customers or write reports in more than one language. Catalonia's multilingual labor market is especially adequate to measure the effect of language knowledge on an individual's occupational selection. Since Catalan language became co-official with Castilian (Spanish) its knowledge has increased substantially among individuals who come from a Catalan-family background as well as among those who do not. This increase, however, has not undermined the intensive use of Castilian in most communication exchanges. Using census data for 1991 and 1996, and controlling for endogeneity of Catalan knowledge, we find that reading and speaking as well as writing Catalan reinforce occupational selection into more communication intensive occupations: entrepreneurial, trade, and service activities and white-collar occupations, both for individuals born in and outside Catalonia. Catalan knowledge also contributes to selection away from temporary employment and into permanent employment. Interestingly, this effect of language knowledge on these occupations is stronger for women than for men.

  Work in Progress

 

 

 

  • On Risk Aversion, Wealth, and On-the-Job Search.working icon
  • Estimating Investment under Liquidity Constraints On and Off Steady State.working icon
  • Applying and Approving a Loan: Needs and Fears working icon 

Joint with Emilia García-Appendini (Università Bocconi).

  • Wage Returns of a Diglossic Language: Evidence from Galicia working icon 

Joint with Jose Maria Da Rocha (Universidade de Vigo).